Ứng dụng mạng Kolmogorov-Arnold trong nhận dạng tĩnh mạch ngón tay
và
Abstract
Recently, Kolmogorov–Arnold Networks (KANs) have gained attention as a novel architecture capable of optimizing nonlinear learning processes and enhancing model performance without significantly increasing computational complexity. This study applies KANs to the biometric recognition task and compares their performance with popular CNN architectures such as InceptionV3, EfficientNet, and MobileNetV3. Through experiments conducted on two biometric datasets, FV_USM and SDUMLA-HMT, the proposed KAN-based model achieves accuracy rates of 99.3% and 96.2%, outperforming traditional architectures. Despite having a higher number of parameters (34.81 million), the KAN model maintains an inference time of 1.0096ms, comparable to InceptionV3 (1.006ms) and significantly faster than EfficientNet_B4 (1.349ms). Furthermore, the model’s computational complexity (539.12 MMAC) demonstrates its feasibility for real-world deployment in biometric recognition systems requiring high accuracy and fast processing speed. These findings highlight the potential of KANs in improving the performance of deep learning models for biometric recognition tasks.
Tóm tắt
Gần đây, mạng Kolmogorov–Arnold (KANs) đã thu hút sự quan tâm như một kiến trúc mới có khả năng tối ưu hóa quá trình học phi tuyến tính, giúp cải thiện hiệu suất mô hình mà không làm tăng đáng kể độ phức tạp tính toán. Trong nghiên cứu này, mạng KANs được ứng dụng vào bài toán nhận dạng tĩnh mạch ngón tay và so sánh với các mô hình CNN phổ biến như InceptionV3, EfficientNet và MobileNetV3. Kết quả thực nghiệm trên hai tập dữ liệu FV_USM và SDUMLA-HMT cho thấy mô hình đề xuất đạt độ chính xác 99,3% và 96,2%, cao hơn so với các kiến trúc truyền thống. Mặc dù có số lượng tham số cao hơn (34,81 triệu), mô hình đề xuất có thời gian tính toán 1,0096ms, tương đương với InceptionV3 (1,006ms) và nhanh hơn đáng kể so với EfficientNet_B4 (1,349ms). Hơn nữa, mức độ phức tạp tính toán của mô hình (539,12 MMAC) cho thấy khả năng triển khai thực tế trong các hệ thống nhận dạng sinh trắc học có yêu cầu cao về độ chính xác và tốc độ xử lý. Những kết quả này khẳng định tiềm năng của mạng KANs cho bài toán nhận dạng sinh trắc học.
Article Details
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
Tài liệu tham khảo
Arnold, V. (1968). On functions of several variables. Russian Mathematical Surveys, 23(4), 49–112.
Asaari, M. S. M., Suandi, S. A., & Rosdi, B. A. (2014). Fusion of band limited phase only correlation and width centroid contour distance for finger based biometrics. Expert Systems with Applications, 41(7), 3367-3382. https://doi.org/10.1016/j.eswa.2013.11.033
Bilal, A., Sun, G., & Mazhar, S. (2021). Finger-vein recognition using a novel enhancement method with convolutional neural network. Journal of the Chinese Institute of Engineers, 44(5), 407-417.
https://doi.org/10.1080/02533839.2021.1919561
Fronitasari, D., & Gunawan, D. (2017). Palm vein recognition by using modified of local binary pattern (LBP) for extraction feature. In 2017 15th international conference on Quality in Research (QiR): International symposium on electrical and computer engineering (pp. 18-22). IEEE. https://doi.org/10.1109/qir.2017.8168444
Gabor, D. (1946). Theory of communication. Journal of the Institution of Electrical Engineers, 93(26), 429–457.
https://doi.org/10.1049/ji-3-2.1946.0076
Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., Courville, A., & Bengio, Y. (2014). Generative adversarial nets. Advances in Neural Information Processing Systems, 27, 2672–2680.
Howard, A., Sandler, M., Chu, G., Chen, L. C., Chen, B., Tan, M., Wang, W., Zhu, Y., Pang, R., Vasudevan, V., Le, Q.V., & Adam, H. (2019). Searching for mobilenetv3. In Proceedings of the IEEE/CVF international conference on computer vision (pp. 1314-1324). https://doi.org/10.1109/iccv.2019.00140
Hu, N., Ma, H., & Zhan, T. (2020). Finger vein biometric verification using block multi-scale uniform local binary pattern features and block two-directional two-dimension principal component analysis. Optik, 208, 163664. https://doi.org/10.1016/j.ijleo.2019.163664
Kolmogorov, A. (1956). On the representation of continuous functions of several variables by superpositions of continuous functions of lesser variable count. In Dokl. Akad. Nauk SSSR (Vol. 108, No. 2).
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2012). ImageNet classification with deep convolutional neural networks. Advances in Neural Information Processing Systems, 25, 1097–1105.
Kuang, H., Zhong, Z., Liu, X., & Ma, X. (2020). Palm vein recognition using convolution neural network based on feature fusion with HOG feature. In 2020 5th International Conference on Smart Grid and Electrical Automation (ICSGEA) (pp. 295-299). IEEE. https://doi.org/10.1109/icsgea51094.2020.00070
Li, C., & Xu, X. (2025). Semi-Supervised Learning with Kolmogorov-Arnold Network for MRI Cardiac Segmentation. IEEE Transactions on Instrumentation and Measurement. https://doi.org/10.1109/tim.2025.3550246
Liu, Z., Wang, Y., Vaidya, S., Ruehle, F., Halverson, J., Soljačić, M., Hou, T.Y., & Tegmark, M. (2024). Kan: Kolmogorov-arnold networks. arXiv preprint arXiv:2404.19756.
Ma, X., Jing, X., Huang, H., Cui, Y., & Mu, J. (2017). Palm vein recognition scheme based on an adaptive Gabor filter. Iet Biometrics, 6(5), 325-333.
https://doi.org/10.1049/iet-bmt.2016.0085
Ponnusamy, V., Sridhar, A., Baalaaji, A., & Sangeetha, M. (2019). A palm vein recognition system based on a support vector machine. IEIE Transactions on Smart Processing & Computing, 8(1), 1-7. https://doi.org/10.5573/ieiespc.2019.8.1.001
Sait, A. R. W., AlBalawi, E., & Nagaraj, R. (2024). Ensemble learning driven Kolmogorov-Arnold Networks-based Lung Cancer classification. PloS one, 19(12), e0313386. https://doi.org/10.1371/journal.pone.0313386
Simonyan, K., & Zisserman, A. (2014). Very deep convolutional networks for large-scale image recognition.
Somvanshi, S., Javed, S. A., Islam, M. M., Pandit, D., & Das, S. (2024). A survey on Kolmogorov-Arnold network.
Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J., & Wojna, Z. (2016). Rethinking the inception architecture for computer vision. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 2818-2826). https://doi.org/10.1109/cvpr.2016.308
Tan, M., & Le, Q. (2019). Efficientnet: Rethinking model scaling for convolutional neural networks. In International conference on machine learning (pp. 6105-6114). PMLR.
Wang, J., Cai, P., Wang, Z., Zhang, H., & Huang, J. (2024). Cest-kan: Kolmogorov-Arnold networks for cest MRI data analysis.
Wang, Y. D., Yan, Q. Y., & Li, K. F. (2011). Hand vein recognition based on multi-scale LBP and wavelet. In 2011 International Conference on Wavelet Analysis and Pattern Recognition (pp. 214-218). IEEE. https://doi.org/10.1109/icwapr.2011.6014480
Yang, H., Fang, P., & Hao, Z. (2020). A gan-based method for generating finger vein dataset. In Proceedings of the 2020 3rd International Conference on Algorithms, Computing and Artificial Intelligence (pp. 1-6). https://doi.org/10.1145/3446132.3446150
Yang, J., Shi, Y., & Yang, J. (2010). Finger-vein recognition based on a bank of Gabor filters. In Computer Vision–ACCV 2009: 9th Asian Conference on Computer Vision (pp. 374-383). Springer Berlin Heidelberg. https://doi.org/10.1007/978-3-642-12307-8_35
Yang, Z., Zhang, J., Luo, X., Lu, Z., & Shen, L. (2025). MedKAN: An Advanced Kolmogorov-Arnold Network for Medical Image Classification.
Yin, Y., Liu, L., & Sun, X. (2011). SDUMLA-HMT: A multimodal biometric database. In Biometric Recognition: 6th Chinese Conference, CCBR 2011, Beijing, China, December 3-4, 2011. Proceedings 6 (pp. 260-268). Springer Berlin Heidelberg.
https://doi.org/10.1007/978-3-642-25449-9_33
Zhang, B., Huang, H., Shen, Y., & Sun, M. (2025). MM-UKAN++: A Novel Kolmogorov-Arnold Network Based U-shaped Network for Ultrasound Image Segmentation. IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control. https://doi.org/10.1109/tuffc.2025.3539262
Zhang, X., & Wang, W. (2020). Finger vein recognition method based on GLCM-HOG and SVM. In 2020 IEEE 3rd International Conference on Information Systems and Computer Aided Education (ICISCAE) (pp. 698-701). IEEE. https://doi.org/10.1109/iciscae51034.2020.9236798
Arnold, V. (1968). On functions of several variables. Russian Mathematical Surveys, 23(4), 49–112.
Asaari, M. S. M., Suandi, S. A., & Rosdi, B. A. (2014). Fusion of band limited phase only correlation and width centroid contour distance for finger based biometrics. Expert Systems with Applications, 41(7), 3367-3382. https://doi.org/10.1016/j.eswa.2013.11.033
Bilal, A., Sun, G., & Mazhar, S. (2021). Finger-vein recognition using a novel enhancement method with convolutional neural network. Journal of the Chinese Institute of Engineers, 44(5), 407-417. https://doi.org/10.1080/02533839.2021.1919561
Fronitasari, D., & Gunawan, D. (2017). Palm vein recognition by using modified of local binary pattern (LBP) for extraction feature. In 2017 15th international conference on Quality in Research (QiR): International symposium on electrical and computer engineering (pp. 18-22). IEEE. https://doi.org/10.1109/qir.2017.8168444
Gabor, D. (1946). Theory of communication. Journal of the Institution of Electrical Engineers, 93(26), 429–457. https://doi.org/10.1049/ji-3-2.1946.0076
Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., Courville, A., & Bengio, Y. (2014). Generative adversarial nets. Advances in Neural Information Processing Systems, 27, 2672–2680.
Howard, A., Sandler, M., Chu, G., Chen, L. C., Chen, B., Tan, M., Wang, W., Zhu, Y., Pang, R., Vasudevan, V., Le, Q.V., & Adam, H. (2019). Searching for mobilenetv3. In Proceedings of the IEEE/CVF international conference on computer vision (pp. 1314-1324). https://doi.org/10.1109/iccv.2019.00140
Hu, N., Ma, H., & Zhan, T. (2020). Finger vein biometric verification using block multi-scale uniform local binary pattern features and block two-directional two-dimension principal component analysis. Optik, 208, 163664. https://doi.org/10.1016/j.ijleo.2019.163664
Kolmogorov, A. (1956). On the representation of continuous functions of several variables by superpositions of continuous functions of lesser variable count. In Dokl. Akad. Nauk SSSR (Vol. 108, No. 2).
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2012). ImageNet classification with deep convolutional neural networks. Advances in Neural Information Processing Systems, 25, 1097–1105.
Kuang, H., Zhong, Z., Liu, X., & Ma, X. (2020). Palm vein recognition using convolution neural network based on feature fusion with HOG feature. In 2020 5th International Conference on Smart Grid and Electrical Automation (ICSGEA) (pp. 295-299). IEEE. https://doi.org/10.1109/icsgea51094.2020.00070
Li, C., & Xu, X. (2025). Semi-Supervised Learning with Kolmogorov-Arnold Network for MRI Cardiac Segmentation. IEEE Transactions on Instrumentation and Measurement. https://doi.org/10.1109/tim.2025.3550246
Liu, Z., Wang, Y., Vaidya, S., Ruehle, F., Halverson, J., Soljačić, M., Hou, T.Y., & Tegmark, M. (2024). Kan: Kolmogorov-arnold networks. arXiv preprint arXiv:2404.19756.
Ma, X., Jing, X., Huang, H., Cui, Y., & Mu, J. (2017). Palm vein recognition scheme based on an adaptive Gabor filter. Iet Biometrics, 6(5), 325-333.
https://doi.org/10.1049/iet-bmt.2016.0085
Ponnusamy, V., Sridhar, A., Baalaaji, A., & Sangeetha, M. (2019). A palm vein recognition system based on a support vector machine. IEIE Transactions on Smart Processing & Computing, 8(1), 1-7. https://doi.org/10.5573/ieiespc.2019.8.1.001
Sait, A. R. W., AlBalawi, E., & Nagaraj, R. (2024). Ensemble learning driven Kolmogorov-Arnold Networks-based Lung Cancer classification. PloS one, 19(12), e0313386. https://doi.org/10.1371/journal.pone.0313386
Simonyan, K., & Zisserman, A. (2014). Very deep convolutional networks for large-scale image recognition.
Somvanshi, S., Javed, S. A., Islam, M. M., Pandit, D., & Das, S. (2024). A survey on Kolmogorov-Arnold network.
Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J., & Wojna, Z. (2016). Rethinking the inception architecture for computer vision. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 2818-2826). https://doi.org/10.1109/cvpr.2016.308
Tan, M., & Le, Q. (2019). Efficientnet: Rethinking model scaling for convolutional neural networks. In International conference on machine learning (pp. 6105-6114). PMLR.
Wang, J., Cai, P., Wang, Z., Zhang, H., & Huang, J. (2024). Cest-kan: Kolmogorov-Arnold networks for cest MRI data analysis.
Wang, Y. D., Yan, Q. Y., & Li, K. F. (2011). Hand vein recognition based on multi-scale LBP and wavelet. In 2011 International Conference on Wavelet Analysis and Pattern Recognition (pp. 214-218). IEEE. https://doi.org/10.1109/icwapr.2011.6014480
Yang, H., Fang, P., & Hao, Z. (2020). A gan-based method for generating finger vein dataset. In Proceedings of the 2020 3rd International Conference on Algorithms, Computing and Artificial Intelligence (pp. 1-6). https://doi.org/10.1145/3446132.3446150
Yang, J., Shi, Y., & Yang, J. (2010). Finger-vein recognition based on a bank of Gabor filters. In Computer Vision–ACCV 2009: 9th Asian Conference on Computer Vision (pp. 374-383). Springer Berlin Heidelberg. https://doi.org/10.1007/978-3-642-12307-8_35
Yang, Z., Zhang, J., Luo, X., Lu, Z., & Shen, L. (2025). MedKAN: An Advanced Kolmogorov-Arnold Network for Medical Image Classification.
Yin, Y., Liu, L., & Sun, X. (2011). SDUMLA-HMT: A multimodal biometric database. In Biometric Recognition: 6th Chinese Conference, CCBR 2011, Beijing, China, December 3-4, 2011. Proceedings 6 (pp. 260-268). Springer Berlin Heidelberg.
https://doi.org/10.1007/978-3-642-25449-9_33
Zhang, B., Huang, H., Shen, Y., & Sun, M. (2025). MM-UKAN++: A Novel Kolmogorov-Arnold Network Based U-shaped Network for Ultrasound Image Segmentation. IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control. https://doi.org/10.1109/tuffc.2025.3539262
Zhang, X., & Wang, W. (2020). Finger vein recognition method based on GLCM-HOG and SVM. In 2020 IEEE 3rd International Conference on Information Systems and Computer Aided Education (ICISCAE) (pp. 698-701). IEEE. https://doi.org/10.1109/iciscae51034.2020.9236798